A Framework for Adaptive E-Learning

download A Framework for Adaptive E-Learning

of 212

  • date post

    20-Feb-2015
  • Category

    Documents

  • view

    140
  • download

    1

Embed Size (px)

Transcript of A Framework for Adaptive E-Learning

A Framework for Adaptive e-Learningby

Keith Maycock, BSc.

Dissertation submitted in partial fulllment of the requirements for candidate for the degree of Doctor of Philosophy

Department of Computer Science, National University of Ireland, Maynooth, Co. Kildare, Ireland. Supervisor: Dr. John G. Keating October 2010

Abstract

Adaptive learning systems attempt to adapt learning content to suit the needs of the learners using the system. Most adaptive techniques, however, are constrained by the pedagogical preference of the author of the system and are always constrained to the system they were developed for and the domain content. This thesis presents a novel method for content adaptation. A personal prole is described that can be used to automatically generate instructional content to suit the pedagogical preference and cognitive ability of a learner in real time. This thesis discusses the manifestation of measurable cognitive traits in an online learning environment and identies cognitive resources, within instructional content, that can be used to stimulate these manifestations.

There exists two main components for the learning component: Content Analyser and a Selection Model. The Content Analyser is used to automatically generate metadata to encapsulate cognitive resources within instructional content. The analyser is designed to bridge the perceived gap found within instructional repositories between inconsistent metadata created for instructional content and multiple metadata standards being used. All instructional content that is analysed is repackaged as Sharable Content Object Reference Model (SCORM) conforming content. The Selection Model uses an evolutionary algorithm to evolve instructional content to a Minimum Expected Learning Experience (MELE) to suit the cognitive ability and pedagogical preference of a learner. The MELE is an approximation to the expected exam result of a learner after a learning experience has taken place. Additionally the thesis investigates the correlation between the cog-

nitive ability and pedagogic preference of an author of instructional content and the cognitive resources used to generate instructional content. Furthermore the eectiveness of the learning component is investigated by analysing the learners increase in performance using the learning component against a typical classroom environment.

3

Acknowledgments

I am very grateful to my wife Bernie who has stuck beside me throughout a number of years, that has seen us live through all kinds of experiences. In particular, when Conor was born and Bernie did all the night feeds to ensure that I nish the PhD. Bernies support throughout the years focused me and helped to nish the PhD after starting a new job, without Bernies support and help I would not have nished in 2010. However with the thanks there is also disappointment, as after successfully defending the thesis I arrived home to be introduced to Dr Manny (our dog!), Dr Mammy, Dr Paul and Dr Conor, by Paul who was nearly three. I need to work at becoming a Professor now.

In addition to the support of my wife and children I have had great support from my family who have always supported my crazy ideas and instilled in me from an early age that anything was achievable, thanks Mam and Dad. I would like to thank my older broth Jonathan who sat behind me in the research lab at NUI, Maynooth. Even though we researched in completely dierent elds the respect for each other grew over the years at Maynooth and our countless discussions helped me to take a more pragmatic view towards my research. Education is a fantastic element of life and I am delighted that my mother and sister are currently pursuing third level qualications. I am so proud that education is held in such high regard with my family and one day we can all be Doctors. I am very grateful to Carl my younger brother, who is one of my best friends for his support over the years. I believe that once Carl nds himself he will excel in whatever eld he applies himself to.

John Keating has been an inspiration and friend throughout the PhD process. His guidance has helped me cope with the stresses of life as well as focusing on constructing a solution for the PhD. Thank you John.

To the guys and girls of the research labs in Maynooth I am eternally grateful for all our conversation explaining the latest trends in technology. In particular, great thanks to Jonathan Lambert who has been a friend and teacher, since rst meeting him in the library at NUIM. Jonathan has been extremely supportive throughout the years and his expert understanding of programming helped me to complete my nal year undergraduate project. Jonathan and I have had, and will continue to have countless conversations investigating pedagogic strategies and technologies and working together now I am sure he will achieve whatever he wants. I am very happy that we have been lucky enough to nd work together.

5

ContentsPage 1 Introduction and Research Question 1.1 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1.1 1.1.2 1.1.3 1.2 1.3 1.4 1.5 Cognitive traits and eduction philosophy . . . . . . . . . . . Content Adaptation using technology . . . . . . . . . . . . . 1 2 3 7

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Thesis Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Outline of the thesis . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Concusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 22

2 Theory and Background 2.1

Adaptive Hypermedia Systems . . . . . . . . . . . . . . . . . . . . . 25 2.1.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

2.2

Sharable Content Object Reference Model . . . . . . . . . . . . . . 27 2.2.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

2.3

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 36

3 Optimal Personal Prole 3.1

Environmental contexts of a learning environment . . . . . . . . . . 37

i

3.1.1 3.2

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

Adaptation independent of domain . . . . . . . . . . . . . . . . . . 39 3.2.1 3.2.2 3.2.3 3.2.4 Multiple Representation Approach . . . . . . . . . . . . . . 40 Exploration Space Control . . . . . . . . . . . . . . . . . . . 41 Critique of adaptive strategies . . . . . . . . . . . . . . . . . 42 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

3.3

Working Memory Capacity . . . . . . . . . . . . . . . . . . . . . . . 44 3.3.1 3.3.2 3.3.3 3.3.4 3.3.5 3.3.6 Baddeley Model . . . . . . . . . . . . . . . . . . . . . . . . . 45 Nelson Cowans Model . . . . . . . . . . . . . . . . . . . . . 46 Ericsson and Kintsch . . . . . . . . . . . . . . . . . . . . . . 47 Trackable Manifestations of WMC . . . . . . . . . . . . . . . 47 Personal Prole model . . . . . . . . . . . . . . . . . . . . . 48 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

3.4

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 54

4 Content Analyser 4.1

Inside the Content Analyser . . . . . . . . . . . . . . . . . . . . . . 55 4.1.1 4.1.2 Developing compatible content for the CA . . . . . . . . . . 57 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

4.2

Stimulating Cognitive Resources . . . . . . . . . . . . . . . . . . . . 60 4.2.1 4.2.2 4.2.3 The importance of structure . . . . . . . . . . . . . . . . . . 64 Controlling the instructional space . . . . . . . . . . . . . . 65 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

4.3

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 71

5 Selection Model 5.1

Suitable searching strategies . . . . . . . . . . . . . . . . . . . . . . 72 5.1.1 Ant colony optimisation . . . . . . . . . . . . . . . . . . . . 74

ii

5.1.2 5.1.3 5.1.4 5.1.5 5.1.6 5.1.7 5.2

Cultural algorithm . . . . . . . . . . . . . . . . . . . . . . . 74 Extremal optimisation . . . . . . . . . . . . . . . . . . . . . 75 Reactive Search Optimisation . . . . . . . . . . . . . . . . . 75 Simulated annealing . . . . . . . . . . . . . . . . . . . . . . 76 Genetic Algorithms . . . . . . . . . . . . . . . . . . . . . . . 76 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

Selection model to automatically generate content . . . . . . . . . . 78 5.2.1 5.2.2 High level protocol for learning component . . . . . . . . . . 82 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

5.3

Genetic Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 5.3.1 5.3.2 Genetic Algorithms explored . . . . . . . . . . . . . . . . . . 83 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

5.4

Using a GA for course construction . . . . . . . . . . . . . . . . . . 87 5.4.1 5.4.2 5.4.3 5.4.4 5.4.5 Comparable problem with complete solution space . . . . . . 88 Genetic Operators for evolving content . . . . . . . . . . . . 91 Avoiding a the crowding problem . . . . . . . . . . . . . . . 100 GA for Optimal Learning Objects . . . . . . . . . . . . . . . 102 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

5.5

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 105

6 Learning Component Environment 6.1

Moodle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 6.1.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107